Regarding neural natural language generation models may propagate conspiracy theories by memorization from Findings of ACL 2021.